[OpenSIPS-Users] refer scenario - record-route header

[Laszlo]

Hello,


Just tried to play with the b2b refer scenario with opensips.
The config is pretty much the default LB config from opensips.org, so
nothing sexy in the conf, it works fine without the b2b stuff. LB
destinations are reachable through the private ip of the server.

if I use the b2b topology hiding scenario, it also work fine.
when I use the refer scenario, things goes mad :)

Sending this to the caller party (default LB config):

U 2013/04/30 02:19:05.920262 opensips_public_ip:5060 -> caller_ip:5060

SIP/2.0 183 Session Progress.

Via: SIP/2.0/UDP
caller_ip:5060;received=caller_ip;branch=z9hG4bK25704719;rport=5060.

Record-Route: <sip:username at opensips_private_ip
;r2=on;lr;ftag=as38e68f66;did=a3c.83df6ae1>.

Record-Route:<sip:username at opensips_public_ip
;r2=on;lr;ftag=as38e68f66;did=a3c.83df6ae1>.



New call  with the refer settings applied:

U 2013/04/30 02:13:50.130487 opensips_public_ip:5060 -> caller_ip:5060
SIP/2.0 183 Session Progress.
Record-Route: <sip:6288808754418@;r2=on;lr;ftag=as439eda7d;did=35.b69971a2>.
Record-Route: <sip:6288808754418 at opensips_public_ip
;r2=on;lr;ftag=as439eda7d;did=35.b69971a2>.


In the first record-route header, the host part is missing.

The relevant lines from the syslog in this case:
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lumps_len: lumps_len called with null send_sock
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:process_lumps: null bind_address
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket
Apr 30 02:13:47 svr2 /usr/local/sbin/opensips[30081]:
CRITICAL:core:lump_check_opt: null send socket


What can cause this?
It is a multihomed enviroment with opensips 1.9.
# opensips -V
version: opensips 1.9.0-notls (x86_64/linux)
flags: STATS: Off, USE_IPV6, USE_TCP, DISABLE_NAGLE, USE_MCAST, SHM_MEM,
SHM_MMAP, PKG_MALLOC, F_MALLOC, FAST_LOCK-ADAPTIVE_WAIT
ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16,
MAX_URI_SIZE 1024, BUF_SIZE 65535
poll method support: poll, epoll_lt, epoll_et, sigio_rt, select.
svnrevision: 2:9980
@(#) $Id: main.c 9790 2013-02-15 10:14:34Z bogdan_iancu $
main.c compiled on 00:52:55 Apr 30 2013 with gcc 4.4.6


-Laszlo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20130429/f71679d5/attachment.htm>